Attempts to detect intermolecular interactions have a long history. Perhaps the best known interaction assay is the yeast 2-hybrid assay which is particularly useful in detecting interactions between proteins. See, for example, Evangelista, C., et al., Trends in Cell Biol. (1996) 6:196-199. This assay, however, requires the presence of a relatively elaborate reporter system and is restricted to interactions which take place in the nucleus, since the detectable reporter is generated by activating transcription.
Alternatively, assays have been developed which rely on the assessment of the activity of a reconstituted enzyme. For example, the cloned enzyme-donor immunoassay (CEDIA®) is described in U.S. Pat. No. 5,643,734. In this assay, an analyte is coupled to a portion of an enzyme designated an “enzyme donor,” and allowed to interact with an “enzyme acceptor” which comprises the remaining portions of the enzyme that are required for activity. In the absence of an interfering analyte binding protein, such as an antibody, this interaction occurs spontaneously and the enzyme reconstitutes itself and exhibits activity. However, when analyte binding protein is present, this reconstitution is prevented. The amount of analyte in solution can then be measured by virtue of the ability of the analyte to compete for the analyte binding protein thus permitting reconstitution of the enzyme. It is seen that this approach relies on the ability of the two portions of the enzyme spontaneously to interact; thus, the mediation of carrier substances is unnecessary and the assay is appropriate both for measuring analyte concentration, and interactions between molecules, i.e., the analyte and analyte binding protein. However, only a limited number of constructs can be made wherein an enzyme donor portion and an enzyme acceptor portion spontaneously recombine. The exemplified enzyme for CEDIA® is β-galactosidase.
Where the components of the enzyme do not spontaneously recombine, the interaction can be adapted to detect interactions between components that are coupled to each portion of the enzyme. Such an approach has been applied using dihydrofolate reductase (DHFR) as an exemplary reporting system by Remy, I., et al., Proc. Natl. Acad. Sci. USA (1999) 96:5394-5399; U.S. Pat. No. 6,270,964. In this system, it is necessary to supply some impetus for the reconstitution of DHFR; coupling of each portion of DHFR to a leucine zipper sequence has been found effective as a model system. This model system can be adapted to detect the interaction of any two substances where the interaction draws their coupled DHFR portions closer together. The interaction is detected, then, by assessing the activity of the enzyme. This system thus avoids the more complex reporter functions characteristic of the yeast 2-hybrid method.
An early embodiment of this type of approach is described by Johnsson, N., et al., Proc. Natl. Acad. Sci. USA (1994) 91:10340-10344. The molecule to be reconstituted in this case is ubiquitin which results in proteins fused thereto in being cleaved by ubiquitin-specific proteases. Because the proteolysis occurs in the proteasome, this method permits monitoring protein/protein interactions as functions of time at the natural sites of their interaction in cells.
None of the foregoing methods provide a direct immediate measurement of the interaction of two individual substances, i.e., one that does not employ a reaction cascade for detection. Neither do the results of prior art methods vary linearly with affinity. The present invention provides such results.